At the moment, the package provides a csv reader that is based on the tabular reader protocol. It is however required to define a way to select the appropriate reader when using the CLI interface when creating snapshots.
Options:
- Infer reader from file extension:
csv -> CvReader, etc. This option seems too limited when implemented alone.
- Let user define reader in the CLI:
dstrack ... --reader <reader_identifier>.
- Let
dstrack discover entry points that match the tabular reader protocol. This might require more effort from users as they have to declare the entry points in their codebase.
At the moment, the package provides a csv reader that is based on the tabular reader protocol. It is however required to define a way to select the appropriate reader when using the CLI interface when creating snapshots.
Options:
csv->CvReader, etc. This option seems too limited when implemented alone.dstrack ... --reader <reader_identifier>.dstrackdiscover entry points that match the tabular reader protocol. This might require more effort from users as they have to declare the entry points in their codebase.